Every month, someone new presents themselves as an “AI expert” ready to transform how audit and finance teams work. Many of them have never signed off an audit file, tested a control, or answered to a regulator. Cynthia Merhej, founder of AIKit LB and a CPA, CISA, and AAIA with a Master’s degree in Artificial Intelligence, says that is exactly where firms go wrong. After training more than a thousand auditors, accountants, and finance professionals across Lebanon and the wider region, she has seen what happens when firms choose their AI enabler on enthusiasm rather than credentials.
We sat down with her to talk about her path from CPA to AI auditor, and what firms should look for before trusting anyone to bring AI into their practice.
Q: You started as a CPA. How did you end up in AI?
A: Through audit, not around it. As a CPA, I was trained to ask where the numbers come from and whether I can rely on them. The CISA took that same question into IT systems and controls. When AI started showing up in client processes and in our own work, the next step was obvious: if a model is producing numbers or conclusions, someone has to be able to audit it. That led me to the Master’s in Artificial Intelligence and then the AAIA. Each step built on the one before. None of them replaced it.
Q: Why do credentials matter so much when a firm chooses someone to lead its AI program?
A: Because AI in audit and finance isn’t a technology project. It’s a professional responsibility. The person guiding your team will shape how staff use AI on client data, what gets documented, and what gets relied on in a file. If they don’t understand professional standards, audit evidence, independence, and confidentiality, they will teach shortcuts that look impressive in a workshop and create problems at review. Credentials aren’t a guarantee, but they tell you the person has been tested on the things that matter when something goes wrong.
Q: What does a generic AI trainer usually miss?
A: The audit trail. A generic trainer will show your staff how to get a fast answer from ChatGPT. They won’t ask whether that output can be documented, reviewed, and defended. In our work, a good answer you can’t evidence is still a weak answer. They also miss professional skepticism. AI sounds confident even when it’s wrong, and auditors need to be trained to challenge it the same way they challenge any other source.
Q: What does the AAIA add on top of the CPA and CISA?
A: The AAIA is ISACA’s Advanced in AI Audit credential, and it’s built for professionals who already hold an audit credential such as the CISA, CIA, or CPA. That says a lot. It doesn’t treat AI as a separate world. It treats it as something auditors must be able to govern, assess, and test. That’s the lens I bring to every program: not just how to use AI, but how to use it in a way you can stand behind.
Q: Why does a Master’s in AI matter if a firm only wants practical training?
A: Because practical training only works when it’s grounded. Understanding how models are built and where they fail is what lets me tell a team, with confidence, which tasks are safe to hand to AI and which ones need a human in the loop. Without that, training becomes a list of tips that’s out of date within six months. With it, people understand why the rules exist, and they keep following them when the tools change.
Q: What should firms ask before hiring an AI program enabler?
A: Three simple questions. Have you worked in audit or finance, and do you hold a professional credential in it? How does your approach handle client confidentiality and documentation? And what happens after the training ends? If the answers are vague, that tells you everything. A good enabler should be able to talk risk with your partners and prompts with your juniors on the same day.
Q: What are the red flags?
A: Promises of instant transformation. A one-size-fits-all course that could be delivered to a marketing agency without changing a slide. No mention of governance at all. And someone who can’t explain where your data goes when staff type it into a tool. If the enabler hasn’t thought about that, your firm ends up carrying the risk.
Q: How does your background shape the way AIKit LB works with firms?
A: Everything we build starts from how audit and finance work is actually done. Our programs are designed around real engagements, real documentation, and real review points, not generic business examples. And I don’t deliver a course and walk away. I stay on afterward, working through the firm’s actual processes until AI is built into daily work with the right controls around it, not just discussed in a meeting room.
Q: Final question. What’s the one thing you want firms to remember when choosing an AI partner?
A: That they’re choosing someone who will influence how their professionals exercise judgment. That deserves the same scrutiny as hiring a senior manager or appointing an external expert. Ask for credentials, ask for experience in your field, and ask how they’ll be accountable for what they teach. The right enabler makes your firm faster and safer at the same time. The wrong one only makes it faster.
The Bottom Line
AI will change how audit and finance work gets done, but the profession’s standards don’t change with it. According to Cynthia Merhej, the safest way to adopt AI is to learn it from someone who has been held to those standards, and who understands both the technology and the responsibility that comes with using it.
About the Author
